Noise generated by computers, routers, NAS, local networks

I would suggest you check out The Hifi podcast with Darren and Duncan. About — The HiFi Podcast with Darren & Duncan

Unlike the Sonos forum-like posters on this thread, Darren Myers has an actual EE degree, awards for his products (analog product of the year by Stereophile), and recently was hired as the VP of Engineering for Parasound.

I realize there is no amount of facts, logic, or experience that can convince anyone on this thread to believe that there is more to sound than what can be measured with a 1KHz test tone.

2 Likes

As an example of what you’re practicing here, let me also just pick from the quoted website whatever fits my opposing point in the argument, like so:

Duncan … is an experienced DIYer when it comes to audio equipment and speakers … working as the Retail Sales Manager of Boulder’s PS Audio … He is a … cable designer … a testing technician for the world’s largest online re-seller of high end audio, The Music Room …

So that makes Duncan a capacity in all things audio without agenda - everyone should follow his recommendations and take his insights for gospel?

For the record, there is more to sound that can be measured, namely a full comprehensive set of measurements - you knew that, right?

3 Likes

This is a strawman argument. I never said trust and certainly not take anything as gospel.
Those are your words. I only encouraged you to listen and make up your own mind. If your qualifications and experience are superior please enlighten us. Or maybe approach it from the perspective that you might learn something new?

There is no “comprehensive” set of measurements that applies to your room, your setup, and your taste in music, or listening volume. Music reproduction is in the early stages of development. Digital even more so.

For the record, when I hear an accomplished engineer say something I disagree with, I don’t treat it as gospel or truth, but it carries more weight than the arguments I’ve read here in this post.

“Every truth has two sides; it is as well to look at both, before we commit ourselves to either.”
-Aesop

Not being a reader of Sonos forums, I don’t know what you mean by that, but it sure sounds like a scattershot ad-hominem attack. I suggest we refrain from attempting to characterize other members of this forum.

On the other hand, characterizing podcasts is another matter. The two folks on the podcast you mentioned both work for, or previously worked for, PS Audio. PS Audio is a company that many of us regard as questionable, due to their long history of questionable products and questionable marketing tactics by their CEO. It’s just a point of view many of us have. And choosing to work for said questionable company seems to many of us like a questionable choice in and of itself. The podcast you reference, by two PS Audio people, would be hard to trust, for many of us. We suspect there might be questionable claims, marketing disguised as pseudo-facts.

3 Likes

Sonos is for users who are value-conscious and looking for a simple and easy-to-use solution. Roon is for those of us who want to eke out the most from our systems and are willing to learn, listen, and be open minded.

I find the ad hominem attacks on Paul and PS Audio distasteful, uninformed, and close-minded. This is coming from someone who has heard but can’t justify their cost. Paul is a pillar of our community and someone that I have a lot of respect for. He’s a lover of roon, and all things audio. His books have taught me a lot about getting the most from my systems. Could you please enlighten me on this “long list of questionable products” that you have personal experience with vs. Amir’s poorly constructed “comprehensive tests”?

The Fox and the Grapes - Aesop

Once upon a time, there was a fox who was very hungry. He was walking through the forest when he saw a beautiful bunch of ripe grapes hanging high on a vine. The fox wanted the grapes very much, but he couldn’t reach them.

The fox tried to jump up and reach the grapes, but he couldn’t reach them. He tried to climb the vine, but it was too high. The fox tried everything he could think of, but he still couldn’t reach the grapes.

Finally, the fox gave up. He was tired and hungry, and he didn’t want to try anymore. He turned and walked away.

As the fox was walking away, he said to himself, “Those grapes were probably sour anyway.”

Moral of the story: It is easy to despise what you cannot get.

1 Like

They are not ad hominem attacks. They are expressions of opinions formed about a company and its founder, long-time CEO, and chief cheerleader. That’s all. Different people may have other opinions. And a disagreement based simply on a different evaluation of things is not an “attack”. I only bring this up to explain why some might not be eager to listen to things coming out of PS Audio.

I wouldn’t speak for all Roon users. I for one was interested in Roon mainly as a digital transport with multi-room capabilities, not to “eke” some magic out of my system, and it didn’t open up my mind to snake oil designed to “fix” the “digital noise” unicorn.

3 Likes

More of a sasquatch, I feel.

1 Like

This is way overdue as I downloaded your file not long after you posted it but you saved me a chunk of change so I thought I’d respond. I finally used my Headphone rig to run your test. I had to set the amp gain to high and run the volume all the way around but there it was, the tone was quiet but clearly there. No hiss, nothing else, just the tone. How did that save me money? I recently received notice that the original EtherRegen was for sale again and I’d given some thought to replacing the switch in that setup with one. As there is no noise at a level far higher than I could ever play actual music I’ve decided to pass on the purchase. Maybe I’ll invest the money in a nice Single Malt which I know from previous experience is a tweak that actually works to make everything sound better. Thanks for the education.

7 Likes

Yeah, that would be more appropriate here in the PNW.

I’d really like to know the way digital streaming works so I can understand how to debunk claims like this.

So there’s a signal that comes to my home via the internet served up by eg Qobuz or from my PC. The signal is decoded by a streamer, fed to a dac, amplified and fed to speakers.

Ignoring the analogue side of the equation, is the problem of fidelity limited to jitter?

If a dac has very low jitter then does that mean it serves up an analogue signal without changing it in the time dimension? That would be a good thing. But what if the signal is already “out of time”? The dac wont make it worse, but has a poor quality streamer/pc already distorted this timing? How does a dac “reclock”?

Please help me get this :see_no_evil::see_no_evil::see_no_evil:

1 Like

A streamer just transfers bits from a network source to a local (to the streamer) device, typically via S/PDIF or USB. As long as it gets the bits right, a modern DAC will take care of all the timing issues. Jitter is not something to worry about.

This is why a $100 Raspberry Pi works just fine as a streamer – all it has to do is pull the bits off the network and send them out one of its ports, transforming the format as it does (a function of software, not the hardware).

No.

Yes.

Since PCs and streamers use buffering when transmitting data, there is no timing problem up until the very last part of the digital chain: the connection to the DAC.

  • If that is USB, there is no re-clocking needed by the DAC, since the USB transfer rate can be controlled by the DAC to closely follow its internal clock.
  • For other types of connection (like S/PDIF or AES), where the transfer rate cannot be controlled, the DAC constantly adjusts its internal clock to make up for the differences between that clock and the data rate. These adjustments are very small, as they are done across thousands of samples, so they are not audible. That is the only “re-clocking” needed in the whole chain.
2 Likes

I’m going to sound like a child here lol. But why?

I have heard that a dac just “does” but can you shed more light on how? The time dimension is probably the most important element in music, so the production of a waveform in time seems important. How is the timing guaranteed?

Jitter is instability of the digital signal in the time domain. The rising and falling edges of the digital 1s and 0s misalign with the clock signal. In extreme cases, it causes sidebands around a given frequency in the analogue output, i.e. it produces signal artefacts in the analogue output of the DAC that don’t exist in the recording. Here’s a typical Jtest output plot

image
Copyright ASR

The centre peak is the 1 kHz signal. The other peaks are artefacts. However they’re at ~ -125 dB relative to signal.

If you were listening to music at the equivalent noise level of a chainsaw, the jitter artefacts are still below the lowest threshold of hearing.

So, as @Bill_Janssen said, jitter is not something to worry about.

2 Likes

Thanks Marian. So usb is almost like the dac is “pulling” the information in the correct timing governed by its internal clock.

Excuse my ignorance but how does it know what it should be receiving at what precise time?

Does that mean that “femto” clocks can do this more accurately?

You also replied that jitter is not the only error possible. I wonder what else can corrupt fidelity in the digital domain?

Thanks for sharing your knowledge!

Typically with an extremely accurate quartz crystal oscillator used as the clock.

For instance, this write-up by dCS.

1 Like

Dave, its great to hear I saved you from a superfluous purchase, though I’m sure Messrs. Crespi and Swenson will be hating on me for having diddled them out of a sale! :rofl:

I wholeheartedly support your suggestion of a Single Malt purchase to heighten your musical enjoyment. And please, share your impressions of your chosen libation and join in the banter:

3 Likes

Almost :slight_smile: The USB controller in the DAC tells the sender to slow down a little or speed up a little, depending on whether the controller’s buffer tends to fill up or empty if the current rate were maintained. These adjustments make sure that the buffer never fills up completely (overflow) or empties completely (underflow) during playback. As long as that doesn’t happen, the DAC can continue consuming samples from that buffer at its own pace.

The only timing information in digital audio is the sample rate, i.e. the number of samples per second. To simplify things a bit, let’s say you play a CD format track, which has a rate of 44100 samples per second, or 44.1 kHz. Say the DAC sets its clock at a frequency exactly 256 times higher than the sample rate; in this case, about 11.3 MHz. As soon as there are enough samples in the controller’s buffer, the DAC starts playback and processes (converts and outputs) the first sample, then processes the next sample after 256 “ticks” of its clock, and then the next sample after another 256 ticks, a.s.o. As long as the next sample is present in the buffer when the time comes to be processed, everyone is happy.

2 Likes

As a matter of fact they tend ti be well informed (can’t speak for all of them, but generally) by an actual understanding of electricity, something that Uncle Paul either lacks or hides very well.

Yes, he definitely is a pillar of some community. Alas, many of us want no more part of it than we’d want to be in the Flat Earth Society or 2+2 is 5 community.

Amir’s test do not pretend to tell whether you will like the sound because this is by definition personal and subjective. They do show very well how close the output if a device is to the ideal original signal (I don’t think he had been ever found cooking the books as it were). For what they are – a limited set of standardized measurements that lets you compare equipment across certain parameters.

I don’t think I have any equipment that is on ASR’s top list (because aesthetics, ergonomics, and personal preferences etc. are also important) but I would say that for people interested in accurate reproduction of music (rather than just lining their own pockets) Amir is far more of a pillar of a community than Uncle Paul could ever hope to be even if he could stop swindling for a moment.

No. Jitter might be a problem at the point where audio signal, clocked by the source, is delivered to the DAC (so it is only a problem if your DAC is fed by S/PDIF or I2S connection from a bad streamer with inaccurate clock). Up to the streamer in your house, audio data is sent in TCP packets with no timing information. Multiple routers, switches, etc. sitting between the streaming service and you receive those packets, quite possibly out of order and at random times, reassemble them, and send them further on their way. Timing information, in a way where jitter could be an issue, appears only at the final step when the audio signal is fed to the DAC. If you are using an all-in-one device, and that device has a somewhat accurate clock, audio data will reach the DAC itself with sufficiently accurate timing. If you are feeding the DAC with USB, data still has no timing until it gets to the DAC, so it’s still not an issue. Only if you send the signal over S/PDIF (slightly better, at least it is designed forv it) or, God forbid, I2S (which isn’t designed for external transmission in the first place) signal is timed by the source and, depending on how good the source clock is, and how appropriate the cable is for the signal, there can, possibly, be some timing issues. And even then it is probably not actually audible, unless something quite spectacularly badly designed.

5 Likes